© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    Natural Language
    Autoencoders
    NEW    Assistant AxisNEWCircuit TracerUPDATESteerSAE EvalsExportsAPI Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 24120
    Prev
    Next
    INDEX
    Explanations

    the letter 's' and variations related to it, suggesting it monitors for plural forms or possessives in the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    616
    +0.12
    0.5%
    497
    +0.11
    0.5%
    362
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    616
    +0.12
    0.05
    28
    +0.11
    0.04
    550
    +0.11
    0.04
    Negative Logits
    epad
    -0.73
    Reference
    -0.67
    eus
    -0.66
    potion
    -0.66
    ertodd
    -0.64
    ording
    -0.64
    cloth
    -0.63
    iator
    -0.60
     assumes
    -0.60
    event
    -0.59
    POSITIVE LOGITS
     largest
    0.99
     borders
    0.95
     own
    0.94
     tallest
    0.93
     poorest
    0.93
     shores
    0.91
     finest
    0.90
     newest
    0.88
     skyline
    0.87
     biggest
    0.86
    Activations Density 0.112%

    No Known Activations